Structure of forecast error covariance in coupled atmosphere–chemistry data assimilation

Abstract

Abstract. In this study, we examined the structure of an ensemble-based coupled atmosphere-chemistry forecast error covariance. The Weather Research and Forecasting (WRF) model coupled with Chemistry (WRF-Chem), a coupled atmosphere-chemistry model, was used to create an ensemble error covariance. The control variable includes both the dynamical and chemistry model variables. A synthetic single observation experiment was designed in order to evaluate the cross-variable components of a coupled error covariance. The results indicate that the coupled error covariance has important cross-variable components that allow a physically meaningful adjustment of all control variables. The additional benefit of the coupled error covariance is that a cross-component impact is allowed; e.g., atmospheric observations can exert an impact on chemistry analysis, and vice versa. Given the realistic structure of ensemble forecast error covariance produced by the WRF-Chem, we anticipate that the ensemble-based coupled atmosphere-chemistry data assimilation will respond similarly to assimilation of real observations.

New Era of Air Quality Monitoring from Space: Geostationary Environment Monitoring Spectrometer (GEMS)

“…An ensemble-based meteorology-chemistry coupled DA system has been developed by interfacing WRF-Chem with the maximum likelihood ensemble filter (MLEF;Županski 2005). Park et al (2015) showed that, in the coupled DA system, the cross-variable components of forecast error covariance made physically meaningful adjustments to all the control variables. They also showed that the coupled error covariance allowed cross-component impacts (e.g., Lim et al 2015;Lee et al 2017).…”

A case study involving single observation experiments performed over snowy Siberia using a coupled atmosphere‐land modelling system

“…The MLEF is an ensemble‐based data assimilation method that includes iterative minimization and is suitable for nonlinear assimilation problems. With respect to coupled systems, the MLEF has been applied to perform coupled chemistry‐atmospheric data assimilation (Lim et al ., ; Park et al ., ).…”

“…The assimilated observations are synthetic pseudo‐observations and are defined as y = x f + σ o (e.g. Park et al ., ). Choosing pseudo‐observations for any variable of interest improves the flexibility of the covariance structure evaluation and allows for a sufficient initial difference between the predicted value and the observation.…”

Impact of Atmospheric and Aerosol Optical Depth Observations on Aerosol Initial Conditions in a strongly-coupled data assimilation system

“…However, unlike variational methods, it has an optimal Hessian preconditioning that not only provides very fast minimization (e.g., 1-2 iterations in most applications) but also a reliable estimate of the analysis error covariance used to define ensemble forecast initial perturbations. MLEF has been applied to high-resolution complex modelling systems that couple clouds, aerosols, carbon, and chemistry (Lokupitiya et al, 2008;Peters-Lidard et al, 2015;Park et al, 2015;Lim et al, 2015;Lee et al, 2017). Mathematical details of the MLEF algorithm can be found in Zupanski et al (2008).…”

“…Assessing its structure in a given application is a good indicator of the potential impact of strongly coupled data assimilation. This is typically achieved by conducting a data assimilation experiment with the assimilation of a single observation (e.g., Thepaut et al 1996;Park et al 2015). Therefore, we conduct an assimilation experiment in which a single AOD observation is assimilated.…”